Overview Figure

Research overview

Research overview diagram for MAI Lab

Research

Three connected areas shape the lab.

Visual Intelligence

Image and video understanding, generation, editing, grounding, detection, segmentation, and robustness under challenging conditions such as low light, rain, and snow.

Physical AI

Perception and understanding that scale into 3D and 4D reconstruction, scene interaction, and embodied reasoning in dynamic physical environments.

Engineering-Driven AI

AI systems for medical imaging, energy systems, mining, and other engineered domains, with an emphasis on robust learning from complex multimodal and sensor-rich data.

Selected Papers

Papers by topic

Visual Generation, Editing, and Personalization

Video Understanding, Detection, and Segmentation